gpt-pp-team-protocol-replay

Fail

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones a comprehensive toolkit from an untrusted external GitHub repository (DanOps-1/gpt-pp-team).
  • [REMOTE_CODE_EXECUTION]: The toolkit contains large, complex scripts like pipeline.py (orchestrator) and card.py (8000 lines of Stripe replay logic) that are executed locally on the host system.
  • [COMMAND_EXECUTION]: The skill uses browser automation frameworks (playwright, camoufox) and network interception tools (mitmproxy) to automate account registration and payment billing flows, which is characteristic of bot-like behavior.
  • [CREDENTIALS_UNSAFE]: The installation and configuration instructions require the user to input extremely sensitive secrets into local JSON files and environment variables, including PayPal account credentials, Cloudflare API tokens, and VLM API keys.
  • [DATA_EXFILTRATION]: The primary goal of the pipeline is to extract and log authentication artifacts such as refresh_token values and protocol capture data (.har files) to local storage, which could be misused if the host is compromised.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 14, 2026, 10:36 AM